Heterogeneous Causal Effects on Well-being and Cooperation in New Zealand
By the end you should be able to
Note
Technical details in appendix
Demographic statistics by religious denomination (NZAVS wave 10, approx~1.4% of NZ adult population).
| Stage | Tool | Purpose |
|---|---|---|
| 1. Target | Causal estimand definition | Binary & shift interventions |
| 2. Learn | SuperLearner (semi-parametric ML) | Outcome & censoring models |
| 3. Adjust | IPC weights + census post-stratification | Population inference |
| 4. Validate | 10-fold cross-validation | Out-of-sample performance |
| 5. Subgroup | Causal forests & policy trees | Detect heterogeneity |
| Estimand | Description | Example |
|---|---|---|
| Binary | All or nothing | 0 vs 1 service/month |
| Shift-up | Increase unless weekly | 0$1, 12, 23, 34 | | Shift-down | Decrease to zero | 40, 30, 20, 1$0 |
Figure 1: ATE: Three-Wave: Binary Intervention
Figure 2: ATE: Three-wave Shift Intervention
Figure 3: ATE: Three-wave Soft Intervention
Figure 4: ATE: Six-wave Soft Intervention GAIN: +1
Figure 5: ATE: Six-wave Soft Intervention LOSS: -1
Figure 6: CATE RATE
Figure 8: ATE: Three-Wave: Binary Intervention
Figure 9: ATE: Three-wave Shift Intervention
Figure 10: ATE: Three-wave Soft Intervention
Figure 11: ATE: Six-wave Soft Intervention WEEKLY/ZER0
Figure 12: ATE: Six-wave Soft Intervention LOSS: -1
Figure 14: Policy Trees: Forgiveness
Figure 15: Policy Trees: Gratitude
Figure 16: Policy Trees for Hlth Fatigue Outcome
Figure 17: Policy Trees: Meaning Sense
Figure 18: Policy Trees: Neighbourhood Community
Figure 19: Policy Trees: Self Esteem
Figure 20: Policy Trees for log Charity Donate Outcome
Note
Click tabs below to explore specific policy trees
This following flowchart shows the decision logic:
START: For each model
|
v
STEP 1: EXCLUSION CHECK
Is RATE QINI < 0 (stat sig) OR RATE AUTOC < 0 (stat sig)?
|
+----+----+
| |
YES NO
| |
v v
EXCLUDED STEP 2: RATE SELECTION CHECK
(Stop) Is RATE QINI > 0 (stat sig) OR RATE AUTOC > 0 (stat sig)?
|
+----+----+
| |
YES NO
| |
v v
SELECTED UNCLEAR
| |
v v
STEP 3: QINI CURVE ANALYSIS
(Applied to all non-excluded models)
| |
v v
Already Check: Is QINI curve positive
selected at any spend level?
|
+----+----+
| |
YES NO
| |
v v
SELECTED UNCLEAR
(final)